Is There an Optimal Size for Magrid Casings?

Discuss the technical details of an "open source" community-driven design of a polywell reactor.

Moderators: tonybarry, MSimon

Aero
Posts: 1200
Joined: Mon Jun 30, 2008 4:36 am
Location: 92111

Post by Aero »

All this talk of various configurations is interesting, but nobody seriously believes that they have a snow balls chance of being built in the near or intermediate term, do they?

I'm doubtful that the dodec will even be built until after the WB-6, WB-7 design shows net power or at least major promise of net power.

Think about it. The navy doesn't seem to be sponsoring research for the sake of research. They have the objective of a power plant in mind it seems to me. They've got a design that might work so I'd expect them to follow the KISS principle and not make any changes that are not needed, at least until after the WB-100 either gives net power or shows a LOT of promise.

As for private capital to build something, I expect they would be even more conservative than the Navy. Maybe once a university backed grant proposal is awarded, uncommon designs might get looked at.
Aero

KitemanSA
Posts: 6114
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

tombo wrote: In you x shaped picture I see 4 three sided vertexes.
One between 1 and 2 another between 2 and 3 another between 3 and 4 another between 4 and 1.
Hmm. Where is the third side, the third coil? I see 4 points with TWO coils each; coils 1&2, 2&3... If I extend it out further, I pick up two more coils. I just don't see a third coil anywhere.

You do realize that the "loop" that appears to be formed by the 4 cross sections is NOT a coil. The current does NOT go around that loop. It enters from two katty-corner leads and exits the two remaining leads. No circuit around, no coil, virtual or otherwise. Each of the 4 points you mention might be considered to have either two or four coils, but not three.
tombo wrote:
You sure have shown that opening small.
Did it on purpose. Tried to minimize. I can easily make it bigger by moving the centers of curvature back. Design algorythm is fairly simple.
tombo wrote:
I thought we were trying to keep the metal out of the way of the cusp region.
Yup. As far as I can tell without analyses, the metal IS out of the way. I guess I'll have to plot up what I think is the mag-field on this holey X. Won't be an analysis, just a graphic representation of my understanding.

KitemanSA
Posts: 6114
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

tombo wrote: Oh, You mean arrayed 4 times around the Z axis.
By George I think he's got it!

Beautiful!!

In the final machine, I think the feed lines may occupy a single pipe at each of the four locations shown rather than two, but I wouldn't swear to that right now. When detailed internal design gets started, we will have to decide on whether to go MPG, 2xMPG or independant coils, but that is for a bit later.
tombo wrote:This looks just about like what I did before except for the extra copies of the feed lines.
And the Holey Xs vice the open corners.
tombo wrote:This is inscribed in a 4m cube with 20" thick coils.
These parameters should be easy to change. (unless the model crashes again.)
Also what is your minimum bend radius? Notice that the square (90 deg bend) and triangle (60 deg bend) have different bend radii.
Yup. It is a result of the angles. The triangle is actually ~2x35.25 or about 70 degrees (vice 60) while the square is 2x54.75 or about 110 degrees (vice 90). Spherical geometry does that to you.

I find that a center of radius at least 3.5 times the width of the pipe suffices for the square bend, though that is for relatively thin pipes. I haven't tried it with thick. When you start getting thick, the differences between spherical and plane geometry start influencing the answer. I suspect that for thick the description algorythm will be based not on width but on projected angle. Anyway, for the triangle, it is closer to 2.5. Any larger results in more seperation.
tombo wrote:I needed that flexibility because in the first version I wanted to make the metal to metal distance the same on the 2 different kinds of kissings. I did not optimize this version for that.
I don't think there is a need to. As long as the narrow dimension meets DrB's seperation criterion, an elongated hole seems fine to me.
tombo wrote:Looks like there will be voids inside the crossovers.
Within the pipes, you caught that. Good, I agree with your assessment completely and have been pretty much planning on it. HOWEVER, it may turn out that the pipe will need to be narrowed in this area because of the weaker field. Exactly how that gets done, if needed, remains to be worked out.

I suspect that shrinking of the TSP around the "Holey X" might suffice. Most of the Holey X will be in partial shadow so will not receive the heat load the rest of the system endures.
tombo wrote:That might give much needed elbowroom for connecting to feed lines etc.
Although the feedlines would need to be repositioned to feed into the areas with voids. (That is easier said than done because of the way the model was put together.)
Let us enjoy the fruits of YOUR labor (thank you VERY much) and leave that til tomorrow. Or maybe Wednesday! :wink:

Edited at 9:29 PM eastern time.
Last edited by KitemanSA on Tue Mar 17, 2009 1:30 am, edited 1 time in total.

KitemanSA
Posts: 6114
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

Aero wrote:All this talk of various configurations is interesting, but nobody seriously believes that they have a snow balls chance of being built in the near or intermediate term, do they?
I am hoping to have a College group (Peninsula College which is a member of the Fusion Club because they have acheived fusion on campus) build a cryo-copper version of the unit that Tom Boydston just posted. The plan is to have a cross section not unlike what I provided about 12 posts back. I hope this can be wound properly on a 15" to 20" sphere. The X-section is about 2.2" in diameter.

It uses square copper tube laid up in a quasi Bitter Magnet manner and should be cooled by either LN, LOX, or the LN2LO2 eutectic.

All told, it should be capable of about 5x the field and at the same field should last much longer. Also, because of improved sphericity, it may produce between 2 and 5 times the output for the same drive.

I have an tentative agreeement with EMC2 to use their tank to run this thing. The only question is whether their power supply is up to the task.

We shall see.

tombo
Posts: 334
Joined: Fri May 09, 2008 1:10 am
Location: Washington USA

Post by tombo »

You do realize that the "loop" that appears to be formed by the 4 cross sections is NOT a coil. The current does NOT go around that loop. It enters from two katty-corner leads and exits the two remaining leads. No circuit around, no coil, virtual or otherwise. Each of the 4 points you mention might be considered to have either two or four coils, but not three.
Yes I do.
They can't be coils because each corner of the square hole is a polyhedron vertex with 3 faces
face 1 is the small rectangular opening
face 2 is the large square coil
face 3 is a triangular coil.
Odd number of faces not even number like the polywell is supposed to have.

There is a lot of flexibility about where to put the feed pipes.
I can keep them in shadow pretty well.
I think having enough of them (perhaps 8 ) spread evenly will help structural strength. Although that is pretty much moot with the cross ties.

So the Peninsula College folks are planning to wind a 2.2" thick bundle of maybe 1/4" copper tubing on a basket ball?
I think I would lay it up as 6 square coils butted up against and between 8 triangular coils.
Are they going to leave the tubing bare to expedite moving on to testing or planning to jacket the whole thing?
I have an tentative agreement with EMC2 to use their tank to run this thing.
Do the people at EMC2 really buy into your corner field funny cusp arguments?
-Tom Boydston-
"If we knew what we were doing, it wouldn’t be called research, would it?" ~Albert Einstein

KitemanSA
Posts: 6114
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

tombo wrote: Yes I do.
They can't be coils because each corner of the square hole is a polyhedron vertex with 3 faces
??? They; which "they" are we talking about?
face 1 is the small rectangular opening
face 2 is the large square coil
face 3 is a triangular coil.
Odd number of faces not even number like the polywell is supposed to have.
Not picturing it. Can you sketch it? I guess I will have to try my hand at a sketch this evening.
What I think you are saying is that the vertex has 3 faces (which for this to be of interest must be COILS), the square coil, the tringular coil, and the NOT-A-COIL coil. But the NOT-A-COIL isn't a coil, so can't be the third coil at the vertex.

There seems to be some confusion about the intent of the even faces issue. Each of the "faces" represents a coil, real or virtual. No coil, no face in the polyhedron of coils. That is why the WB6 configuration is not optimum. The round coils make a span between them that looks much like a line. Eventually they seperate enough to join with the virtual triangle coil. At that point, there is a vertex of 3 coils, the two round and the one virtual triangle. But all faces in that example represent a coil. No coil, no face on the vertex.
So the Peninsula College folks are planning to wind a 2.2" thick bundle of maybe 1/4" copper tubing on a basket ball?
Things are in the preliminary discussion stage. I haven't even had the chance to inform Jeff that DrN tentatively agreed...
This may be beyond them. Their resorces are quite limited, and it is a two year jr college, so students don't stick around as long as one would like.
I have an tentative agreement with EMC2 to use their tank to run this thing.
Do the people at EMC2 really buy into your corner field funny cusp arguments?
No idea. I don't think it is MY idea but DrB's. Indeed, the term came from a reviewer, not DrB.

tombo
Posts: 334
Joined: Fri May 09, 2008 1:10 am
Location: Washington USA

Post by tombo »

they==the "not-a-coil"

The rectangular "not-a-coil" IS a face because it has conductors on its edges.
It touches the 4 new vertices that it creates as it truncates the corner (4 sided vertex) of the (rectified) truncube.
Yes, it can't be a coil precisely because of the geometry.

Image
There seems to be some confusion about the intent of the even faces issue. Each of the "faces" represents a coil, real or virtual. No coil, no face [I don't think so.] in the polyhedron of coils. That is why the WB6 configuration is not optimum. [Agree] The round coils make a span between them that looks much like a line.[actually more like a pair of parallel lines with opposite direction currents. ie. the origin of the rectangular region under discussion] Eventually they seperate enough to join with the virtual triangle coil. At that point, there is a vertex of 3 coils[This is why the cube is truncated to make it into a truncube actually a rectified cube. This is done to make all the vertices even as Dr. B insists they must be.], the two round and the one virtual triangle. But all faces in that example represent a coil. [but I still don't buy this-->] No coil, no face on the vertex.
If you only partially truncate the cube corner and don't fully rectify it you wind up with yes a triangular (virtual) coil and 3 octagonal coils (the squares with the corners trimmed). And, the edges shared by the octagonal coils each have currents flowing in BOTH directions. These currents would cancel out. This kind of thing is exactly what Dr. B was driving at in the geometry portion of the 1989 patent when he insisted on even vertices.
This is why the cube needs to be fully rectified. Furthermore when you truncate that shape again you create odd vertices.

Looky here
The 1989 Patent:
http://www.geocities.com/davidmdelaney/ ... ticles.pdf
P.17 line 60 is where he insists on even vertices.

He never shows round coils and usually refers to currents on polygon edges.

Of note:
He seems to like the Octahedron just fine in spite of the non-opposing field directions of the opposite faces. (this in response to a different argument by someone else.) You mentioned that you like the octahedron.
I also like it because I have a design based on it that is relatively strong and easy to fabricate. Someday I will revisit it to clear up the one objection that worries me.
No idea. I don't think it is MY idea but DrB's.
Can you find the source?
I would feel more comfortable putting a lot of work into this design if someone with more weight than you or me could be shown to support your line of reasoning about the seriousness of the containment breach caused by the X shaped field null.
I think this is really more to the point than all this discussion about geometry.
-Tom Boydston-
"If we knew what we were doing, it wouldn’t be called research, would it?" ~Albert Einstein

Billy Catringer
Posts: 221
Joined: Mon Feb 02, 2009 2:32 pm
Location: Texas

Post by Billy Catringer »

tombo wrote: Of note:
He seems to like the Octahedron just fine in spite of the non-opposing field directions of the opposite faces. (this in response to a different argument by someone else.) You mentioned that you like the octahedron.
I also like it because I have a design based on it that is relatively strong and easy to fabricate. Someday I will revisit it to clear up the one objection that worries me.

Tombo, I am not your boss, so this is just opinion, but if you have what you believe to be a good start on this proceed with it. We definitely need an alternative design to the 6-toroids in a sumo match version that I am working on.

KitemanSA
Posts: 6114
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

tombo wrote:they==the "not-a-coil"

The rectangular "not-a-coil" IS a face because it has conductors on its edges.
It touches the 4 new vertices that it creates as it truncates the corner (4 sided vertex) of the (rectified) truncube.
Yes, it can't be a coil precisely because of the geometry.
Again, we seem to have a difference of understanding. To me, it is the magnetic field arrangement that dictates the number of "faces". This is how I envision it. I am in the process of putting together a very limited spreadsheet to calculate the fields using the Biot-Savart Law. If I get my copy of MATLAB done first, maybe I'll use Indreks m-files and do it more efficiently.
Image
In the left sketch, the magnetic field approaches the upper left conductor from the point cusp in the square coil, turns the corner while crossing the single conductor, and recedes toward the point cusp in the triangle coil. In the right sketch, the magnetic field approaches the upper left conductor and X-Cusp (my new term for the modified “Funny Cups” from the point cusp in the square coil, and either turns the corner while crossing the single conductor, or crosses the square coil corner, turns the corner, crosses the triangle corner, and recedes toward the point cusp in the triangle coil. In either case, what the electron sees is identical, EXCEPT there is no metal in the little tiny hole in the cusp.

Now I do admit that I expect there to be a VERY short segment of line cusp in the very middle of this X-Cusp, but what the hey!

Please note that in his Valencia paper, DrB discusses the gap and the cusp and the need to have the cross connection but that said cross connection had to be magnetically isolated. Effectively that is what I have here; the square coil corner cusp with two magnetically isolated cross connects. As far as I can tell, this meets every requirement set down by DrB in that paper.

hanelyp
Posts: 2255
Joined: Fri Oct 26, 2007 8:50 pm

Post by hanelyp »

Kiteman, the problem with the square split corner you propose is that is doesn't give the same magnetic field. Each of the 4 branches around the square generates it's own magnetic field dominating the local conditions. Try plotting the field ignoring the contribution from the arms feeding the corners of the square. So you end up with cusps running corner to opposite scorner on the square. And not just radial field cusps, but zero field cusps.

KitemanSA
Posts: 6114
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

hanelyp wrote:Kiteman, the problem with the square split corner you propose is that is doesn't give the same magnetic field. Each of the 4 branches around the square generates it's own magnetic field dominating the local conditions. Try plotting the field ignoring the contribution from the arms feeding the corners of the square. So you end up with cusps running corner to opposite scorner on the square. And not just radial field cusps, but zero field cusps.
If you limit your view to the mid-plane, (and ONLY the mid-plane) that may be the case, but before the electrons reach the mid-plane they must pass thru a region that looks like mine. This will reflect all but a small number of the electrons which will continue on and see the very short line cusp.
Please note that the conductors in each corner have current flowing in the same direction. Thus at a distance it looks to be one conductor half way between them. Given the sizes of these things, the wiffleball face (inside the MaGrid) should have enough standoff from the centerline of the MaGrid to make the two "incoming" conductors in each corner of the right-hand sketch above look like a single one, resulting by superposition in a field effectively identical to what would be produced by the straight conductors of the left-hand sketch above.

{{Dark red added by edit}}

tombo
Posts: 334
Joined: Fri May 09, 2008 1:10 am
Location: Washington USA

Post by tombo »

OK, Here is how we keep the zero field cusp region localized radially.
Thanks Kiteman for pushing me to look 3D.

Assuming the field goes in through the main square coils, the field goes outward at locations between the zero field point and the wall and inward at locations between it and center of polywell).
I see the zero field region as an octahedral star. Which is the X from above with tails leading toward the center and away from the center of the polywell.
It is located at the middle of the edge of the cube at the radius of the polyhedron from the polywell center.
There is no zero-field path from the inside to the outside.

It also provides a point cusp aligned with each face and each corner. (in series with the normal ones)
It is based on truncated cube.
The polyhedron is at the midline where the 2 different coils touch.
Think Bitter coils.
Pardon not showing all the coils and there are many small glitches in the model.
The coil taper angle is shown optimized for thermal.
Angle could be optimized for magnetic field or other stuff.
There are several coil shape permutations. For example, they could maintain their planform over their radial length.
Conforming to the field lines will take some serious FEA modeling.

Corner View
Image

Face View (some coils not shown)
Image
Edge View at “funny cusp”
Image

Radial view showing coils lined up with center of polywell.
http://i299.photobucket.com/albums/mm31 ... etail2.jpg

Compare gaps between square coils to gaps between triangular coils. They are similar.
http://i299.photobucket.com/albums/mm31 ... etail1.jpg

EDIT: This will not work because the set of coils closer to the center distorts the spherical shape of the wiffleball.
Last edited by tombo on Sat Apr 04, 2009 7:48 am, edited 1 time in total.
-Tom Boydston-
"If we knew what we were doing, it wouldn’t be called research, would it?" ~Albert Einstein

Billy Catringer
Posts: 221
Joined: Mon Feb 02, 2009 2:32 pm
Location: Texas

Post by Billy Catringer »

That's first rate work, Tombo. Bravo! Now let's see what the physicists think.

KitemanSA
Posts: 6114
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

Tombo,
Geez I wish I could do graphics like that. It sure is an interesting picture. But it was not what I meant.

The discussion we were having (you, me, and hanelyp) had to do with the FIELD produced by my X cusp and the funny cusp, and you and hanelyp had trouble seeing that they gave effectively the same field. I agreed with you provided you were looking ONLY at the field at the mid-plane of the magnets. I suggested you think about the field off plane. Well, I finally got my Excel spreadsheet to work and by plug-and-chug got the results plotted. Here they are. Here you see the two units, my X cusp on the left and the "funny cusp" setup on the right (half of the symmetric systems each). The results are plotted for a plane that is 15 units into the page understanding that the standoff from the line of symmetry to my broad curve nearpoint is 13 units. No they are not I-{clap}-DENTICLE.. (apologies to My Cousin..) but they are dang close. The presence of the rectangular "not-a-coil" plane in the middle does not adversly effect the "four fields at a vertex" requirement at all.

Image
So you see that whatever the field nulls at the magnet's midplane are, the electrons will never get to them except in the very short slot in the center. And at that point there is NO METAL.
So we have a VERY short consrained bit of line-cusp that looks for all practical purposes like a funny cusp (or a point cusp) (its about the same size) that I call an X-cusp and will limit the electrons basically as well as a point cusp and will have NO path to the metal.
Last edited by KitemanSA on Fri Apr 03, 2009 12:18 am, edited 1 time in total.

hanelyp
Posts: 2255
Joined: Fri Oct 26, 2007 8:50 pm

Post by hanelyp »

KitemanSA, could you adjust your calculations to get field strength and direction exactly on the lines dividing the square "not a coil" in half? And especially dead center of the face? As I expected, the field close to the center is looking pretty weak in your plot. Calculation of the same plane for WB-6/7 style kissing coils would also be of interest.

Could you post your spreadsheet someplace for us to experiment with?

Post Reply